Oxide Dispersion Strengthened Nickel Based Alloys via ...

Boise State UniversityScholarWorksMaterials Science and Engineering FacultyPublications and Presentations Department of Materials Science and Engineering

4-10-2015

Oxide Dispersion Strengthened Nickel BasedAlloys via Spark Plasma SinteringSomayeh PasebaniUniversity of Idaho

Aniket K. DuttUniversity of North Texas

Jatuporn BurnsBoise State University

Indrajit CharitUniversity of Idaho

Rajiv S. MishraUniversity of North Texas

NOTICE: this is the author’s version of a work that was accepted for publication in Materials Science and Engineering: A. Changes resulting from thepublishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in thisdocument. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published inMaterials Science and Engineering: A, Volume 630 (2015). doi: 10.1016/j.msea.2015.01.066

Publication InformationPasebani, Somayeh; Dutt, Aniket K.; Burns, Jatuporn; Charit, Indrajit; and Mishra, Rajiv S.. (2015). "Oxide Dispersion StrengthenedNickel Based Alloys Via Spark Plasma Sintering". Materials Science and Engineering: A, 630, 155-169. http://dx.doi.org/10.1016/j.msea.2015.01.066

http://scholarworks.boisestate.edu

http://scholarworks.boisestate.edu/mse_facpubs

http://scholarworks.boisestate.edu/mse_facpubs

http://scholarworks.boisestate.edu/mse

http://dx.doi.org/10.1016/j.msea.2015.01.066



1

Oxide Dispersion Strengthened Nickel Based Alloys via Spark Plasma Sintering

Somayeh Pasebania, Aniket K. Duttb, Jatuporn Burnsc, Indrajit Charita,1 and Rajiv S. Mishrab

a University of Idaho, Moscow, ID 83844-3024, USA b University of North Texas, Denton, TX 76203, USA c Boise State University, Center for Advanced Energy Studies, Idaho Falls, ID 83401, USA

Abstract

Oxide dispersion strengthened (ODS) nickel based alloys were developed via mechanical

milling and spark plasma sintering (SPS) of Ni-20Cr powder with additional dispersion of 1.2 wt.%

Y2O3 powder. Furthermore, 5 wt.% Al2O3 was added to Ni-20Cr-1.2Y2O3 to provide composite

strengthening in the ODS alloy. The effects of milling times, sintering temperature, and sintering

dwell time were investigated on both mechanical properties and microstructural evolution. A high

number of annealing twins was observed in the sintered microstructure for all the milling times.

However, longer milling time contributed to improved hardness and narrower twin width in the

consolidated alloys. Higher sintering temperature led to higher fraction of recrystallized grains,

improved density and hardness. Adding 1.2 wt.% Y2O3 to Ni-20Cr matrix significantly reduced the

grain size due to dispersion strengthening effect of Y2O3 particles in controlling the grain boundary

mobility and recrystallization phenomena. The strengthening mechanisms at room temperature were

quantified based on both experimental and analytical calculations with a good agreement. A high

compression yield stress obtained at 800 °C for Ni-20Cr-1.2Y2O3-5Al2O3 alloy was attributed to a

combined effect of dispersion and composite strengthening.

Keywords: Ni-Cr based alloys; High energy ball milling; Spark plasma sintering; Dispersion

strengthening; Composite strengthening

1. Introduction

Increasing the operating temperatures in coal-fired power plants, gas turbine inlets, and

other high temperature structural components in order to improve their efficiency and economy will

require new materials with high mechanical and creep strength, oxidation and corrosion resistance.

Nickel based alloys are promising candidates for such applications due to their excellent corrosion

resistance at elevated temperatures [1-3].

While conventional nickel based alloys may not be very stable at high temperatures due to

coarsening or dissolution of the second phase particles, nickel based oxide dispersion strengthened

1 Corresponding author. Tel. +1 208 885 5964; Fax: +1 208 885 7462 E-mail address: [email protected] (I. Charit)

This is an author-produced, peer-reviewed version of this article. The final, definitive version of this document can be found online atMaterials Science and Engineering: A, published by Elsevier. Copyright restrictions may apply. doi: 10.1016/j.msea.2015.01.066

2

ODS alloys, reinforced by homogeneously dispersed nanoparticles (usually Y2O3), are quite stable

during high temperature applications in excess of 1000 °C [4]. Homogeneous dispersion of

nanometric stable oxide particles in the matrix of nickel based ODS alloys can act as effective

barriers against dislocation motion [2, 5-7] and improve high temperature mechanical properties

including creep strength [8]. The pinning effects of oxide nanoparticles depend on the mean particle

separation (the mean distance between the particles) which is a direct result of the particle number

density [1]. According to theoretical calculations and experiments, a combination of a mean particle

separation of 100-250 nm for 10-20 nm yttria particles and grain aspect ratio of a minimum of 10

could be promising for high- temperature applications [3, 6].

In nickel based ODS powder containing both Al and Y2O3, different Y-Al-O particles such

as Y3Al5O12, YAlO3 (perovskite), Y4Al2O9 and YAlO3 (hexagonal) can be formed during

consolidation [1]. Recently, it has been noted that adding some minor elements such as Ti and Hf

can replace the Y-Al-O particles with Y-Ti-Hf-O particles. The effects of adding minor elements

such as Ti, Mg, Zr, Ca and Hf to Ni-0.5Al-1Y2O3 (wt.%) was studied by Tang et al. [1], and Hf was

found to be the most effective oxide at refining the formed oxide particles, especially at a

concentration of 0.8 wt.%. Formation of Y2Hf2O7 was found to be responsible for oxide particle

refinement and consequent improvement in mechanical properties through operation of the Orowan

mechanism [9].

Another strengthening mechanism to consider for developing nickel based ODS alloys

would be composite strengthening mechanism or load transfer mechanism [10]. For example,

studies have also shown that submicron Al2O3 of 0.5-1 μm diameter could be efficient for

composite strengthening due to lower density and higher modulus of elasticity [11, 12]. Hornbogen

and Starke [13], and Rosler and Baker [14] predicted that a combination of nanoparticles and

coarser particles dispersed in the microstructure would offer both dispersion strengthening and

composite strengthening. Through dispersion strengthening and composite strengthening as

dominant mechanisms at high temperatures, enhanced mechanical properties would be achieved.

Nickel based ODS alloys are conventionally produced by mechanical alloying (MA) or ball

milling of elemental or pre-alloyed powders in combination with nano-sized Y2O3 (yttria) powder

followed by canning, degassing and consolidation either via hot extrusion (for rods and wires) or

hot isostatic pressing (HIP) and rolling (for sheets) [3].

One of the critical steps in producing nickel based ODS alloys is the milling process in

which powder blends of yttria and pure nickel or pre-alloyed nickel (for example, Ni-20Cr) are

milled, and a fine uniform distribution of yttria particles in the metal matrix can be attained. If

powder blends of yttria and two or more metal powders are milled in addition to homogeneous

yttria dispersion, formation of solid solution may be also achieved [3]. During milling, the metal


3

powder particles become trapped between the colliding balls (milling media) and are cold welded

together while the oxide particles become progressively finer until trapped between layers of metal

powders sandwich forming a composite. After cold welding and particle agglomeration, a fracture

stage occurs and large composite powders break down until a steady state situation is reached

between cold welding and fracturing. Consequently, a uniform distribution of oxide nanoparticles

within metallic components would be achieved [15, 16].

In conventional consolidation methods such as extrusion or HIP, a final annealing at high

temperature is usually required to develop a stable coarse grain structure [16]. The numbers of

thermal processing steps could be eliminated if a pulsed direct current (DC) is simultaneously used

with a uniaxial pressure to primarily sinter the powders [17, 18]. This could reduce the time, cost

and possibly deformation texture in the consolidated materials [19]. Field activated sintering

technique (FAST), also known as spark plasma sintering (SPS) or pulsed electric current sintering

(PECS), applies a pulsed DC to enhance sintering rate of the powders to near full density at

relatively lower temperatures. Pulsed DC flows through the die and powder compact producing heat

via Joule heating mechanism, providing a much higher heating rate and shorter sintering time

compared to conventional sintering techniques. Thus, grain growth during sintering can be

essentially minimized, leading to improvement in mechanical properties [18, 20-22].

There are limited reported applications of SPS in the processing of nickel based ODS alloys

containing Y2O3 as dispersoids [1, 2, 4, 23, 24]. Park et al. [2] developed Ni-22Cr-11Fe-1TiO2, Ni-

22Cr-11Fe-1Y2O3 and Ni-22Cr-11Fe-0.5TiO2-0.5Y2O3 (wt.%) by milling for 40 h in a planetary

ball mill and SPSed the ball milled powder at 1100 °C for 5 min under a pressure of 40 MPa. They

suggested that nano-sized TiO2 and Y2O3 particles dissolved during MA, and then precipitated out

during SPS, forming Y-Ti-O particles. However, Ni-22Cr-11Fe-1Y2O3 exhibited the best

mechanical properties among all of the developed alloys.

In the present study, Ni-20Cr-1.2Y2O3 (wt.%) alloy was processed by ball milling and SPS,

and the effects of milling time and sintering parameters on the properties of sintered nickel based

ODS alloy were investigated. There have been very limited studies on the effects of milling on the

microstructural evolution during milling of nickel based powder [25]. Lopez et al. [25] milled

elemental Ni and Cr powders for 30 h to obtain a nanostructured Ni-20Cr alloy. Such solid

solutionizing reaction occurred mainly due to chemical-heterogeneity-driven diffusion through

interfaces, subgrain boundaries and dislocation cores. In this study, Ni-20Cr, and Ni-20Cr-1.2Y2O3-

5Al2O3 (wt.%) alloys were processed by milling and SPS, and their mechanical properties and

microstructural evolution were studied in detail. Here, 1.2Y2O3 wt.% (or 2 vol.%) and 5Al2O3 wt.%

(or 10 vol.%) were added to Ni-20Cr matrix for dispersion strengthening and composite

strengthening, respectively.


4

2. Experimental

2.1. Powder processing and characterization

Gas atomized Ni-20Cr powder with nominal composition of Ni-19.6Cr-0.2Fe-0.8Mn-0.9Si

(wt.%) and mean particle size of 23.6±1.1 μm; yttrium oxide (yttria/Y2O3) powder with high purity

(99.99%) and mean particle size of 30-40 nm; and aluminum oxide (alumina/ Al2O3) powder with

99.99% purity and mean particle size of 300-400 nm were all procured from the American Elements

Inc.

Powder batches were prepared in a glove box under high purity argon atmosphere and

poured into hardened steel grinding vial (Spex 8001). In order to minimize powder agglomeration

and cold welding during milling, 1 wt.% stearic acid was added to the powder mix prior to the ball

milling process as a process control agent (PCA). The ball milling was carried out in a Spex 8000M

shaker mixer/mill using steel balls 5 mm in diameter and a ball to powder ratio (BPR) of 10:1 (the

powder mass and the ball mass of each batch was 10 g and 100 g, respectively). A variety of Ni-

based alloys altering in milling time (0 h, 2 h and 4 h) and nominal composition (Ni-20Cr, Ni-20Cr-

1.2Y2O3, Ni-20Cr-1.2Y2O3-5Al2O3, wt.%) were milled.

For preparing the Ni-20Cr-1.2Y2O3-5Al2O3 powder, Ni-20Cr-1.2Y2O3 was first milled for 2

h, and then 5 wt.% Al2O3 powder was added to the milled Ni-20Cr-1.2Y2O3 alloy and subsequently

ball milled to distribute all the Al2O3 particles homogeneously. In our former experiments, Al2O3

powder was only blended (i.e. milled without the steel balls) with the milled Ni-20Cr-1.2Y2O3

powder, and the results were unsatisfactory because all the Al2O3 powder particles were found to be

mostly located on the prior particle boundaries after consolidation.

X-ray diffraction (XRD) experiments of the as-milled powders were performed using a

Siemens 5000D diffractometer with Cu-K radiation. Modifications such as k 2 Rachinger and

background correction by Sonnerveld were applied to XRD patterns using the Powder-X software

[26]. Lattice parameters, crystallite size and lattice strain were calculated based on the Nelson-Riley

extrapolation [27] and Williamson-Hall (W-H) formula, respectively [28]. For the instrumental

broadening correction, a fully annealed/unmilled Ni-20Cr powder sample was used as a standard.

The morphology and size distribution of the as-received powder batches and as-milled

powder were analyzed using a Zeiss Supra 35 field-emission gun scanning electron microscope

(FEG-SEM). The milled powders were hot mounted in phenolic powder and polished to 0.05 μm.

The cross section of the hot mounted and polished milled powders were observed in backscatter

electron (BSE) mode in SEM. A SEM micrograph obtained from the as received Al2O3 powder is

shown in Fig. 1 and a high angle annular dark field (HAADF) scanning transmission electron

microscopy (STEM) micrograph is presented in Fig. 1b.


5

2.2. Spark plasma sintering

The ball milled powder was consolidated via SPS using a Dr. Sinter Lab SPS-515S machine

(SPS Syntex Inc., Kanagawa, Japan) with maximum capacity of 30 kN and 1500 A. A Tri-Gemini

cylindrical graphite die with an inner diameter of 12.7 mm and an outer diameter of 38 mm were

used. The inner surface of the die and radial surfaces of punches were covered with a graphite foil

(0.25 mm in thickness) to facilitate the removal of the sintered specimens. In order to inhibit the

diffusion of carbon from the punches and graphite foil to the powder mixture, a thin niobium foil

(0.06 mm in thickness) was placed between the powder and the graphite foils. The die was wrapped

in graphite felt (4 mm in thickness) to minimize heat loss by thermal radiation.

All the SPS experiments were performed under vacuum (7×10-3 Torr or 0.9 Pa), using a

heating rate of 100 oC/min and force of 10 kN (equals to about 80 MPa considering punch

configuration used in this study). An intermediate 15 min dwell time at 450 o C (with 4.5 kN applied

force) was given for all the SPS runs to allow the stearic acid to volatilize. Following that, the

temperature was ramped up to different levels (600 oC, 900 oC, 1000 oC and 1100 oC), kept at those

temperatures for different times (5 and 30 min), and cooled down with at the rate of about 50 oC/min. The temperature was monitored by a K-type thermocouple that was inserted into a hole in

the die such that the tip was located 6 mm away from the sintering powder. The final product was in

the form of a disk with dimensions of 12.5 mm in diameter and 8 mm in thickness. All the Ni-based

alloys processed in this study varying in milling times, compositions, sintering temperatures and

times are listed in Table 1.

2.3. Density and microstructural characterization

Upon SPS, physical density of the bulk specimens was measured using Archimedes’

principle with at least six measurements for each specimen. The final relative density was

calculated as the ratio between the measured density and the theoretical density of each

composition. Electron backscattered diffraction (EBSD) study was performed using a JEOL JSM-

6610LV scanning electron microscope (SEM) equipped with an EDAX/TSL Hikari EBSD system.

The transmission electron microscopy (TEM) specimens were mechanically thinned and

electropolished at a temperature of about -35 oC using a solution of nitric acid and methanol (10:90,

vol.%) and a Fischione twin-jet polisher operating at 30 V. Microstructural studies were conducted

using a JEOL-2010 TEM operating at 200 kV. A focused ion beam (FIB) was used to prepare a

TEM foil from the specimen with a composition of Ni-20Cr-1.2Y2O3-5Al2O3 due to the

unsatisfactory results after electropolishing (likely caused by low conductivity of this alloy). FIB

experiment was done by using an FEI Quanta 3D FEG instrument with a Ga-ion source.

This is an author-produced, peer-reviewed version of this article. The final, definitive version of this document can be found online at Materials Science and Engineering: A, published by Elsevier. Copyright restrictions may apply. doi: 10.1016/j.msea.2015.01.066

6

2.4. Mechanical characterization

The sample surface for microhardness testing was mechanically polished using standard

metallographic procedures involving grinding and polishing down to 0.5 μm finish. The Vickers

microhardness tests were performed with a Leco LM100 microhardness tester at 0.5 kgf (5 N) load

applied for 15 s on the sintered samples. The microhardness tests were repeated on random spots in

the center of each specimen up to 10 times.

Compression testing was performed on an Instron 5982 machine at both 25 °C and 800 oC

applying a strain rate of 10-3 s-1. The disk shaped SPSed specimens were electro-discharge

machined to a square-based rectangular prism shape with dimensions of 4×4×6 mm. After

compression testing, the deformed sample was cooled in water to maintain the microstructure for

further studies planned for our future work.

3. Results

3.1. Effect of milling time

The powder morphology of the Ni-20Cr-1.2Y2O3 powder milled for 0 h (alloy A), 2 h (alloy

B) and 4 h (alloy C) are shown in the secondary electron (SE) SEM micrographs in Figs. 2a, 2c and

2e, respectively. The BSE micrographs from the cross sectional view of powders milled for 0 h, 2 h

and 4 h, mounted and polished are presented in Figs. 2b, 2d and 2f, respectively.

The morphology of Ni-20Cr-1.2Y2O3 powder was a mixture of spherical and irregular

particles with a mean diameter of 23.6±2.1 μm uniformly covered with stearic acid and Y2O3

nanopowder as shown in Fig. 2a. The areas with darker contrast in Fig. 2b were due to stearic acid.

The morphology of Ni-20Cr-1.2Y2O3 powder after 2 h milling is shown in Fig. 2c and exhibited

major agglomeration due to ductile nature of powder particles and high tendency of cold welding at

an early stage of milling.

At an early stage of milling (2 h), the ductile Ni-20Cr powder particles became flat and cold

welded, and trapping of Y2O3 powder particles between Ni-20Cr lamellae likely occurred. The

morphology of powder particles was round and the mean particle size after 2 h milling was

estimated to be 33.6±1.5 μm. The plastic flow and deformed layers in a powder lamella after 2 h of

milling can be clearly discerned in Fig. 2d. Powder agglomeration continued to occur up to 4 h as

shown in Fig. 2e; however, numerous cracks on the powder particles were detected, too.

Accumulation of work hardening led to fatigue and fragmentation within the powder flakes that

could refine powder particle upon further milling. The powder shape was round with a mean

particle size of 39.4±3.1 μm. Figure 2f revealed significant work hardening, cracks and signs of

fragmentation after 4 h of milling.


7

The XRD patterns for Ni-20Cr-1.2Y2O3 powder after 0 h, 2 h and 4 h milling (alloys A, B

and C, respectively) are plotted in Fig. 3. In the XRD pattern of alloy A, the peaks of (111), (200)

and (222) confirmed an fcc crystal structure in Ni-20Cr and the remaining peaks presented a cubic

crystal structure in Y2O3. The peaks in the XRD of Y2O3 were indexed as (222), (400), (411), (422)

and (622) in the order of appearance. The Ni-20Cr peaks exhibited significant broadening after 2 h

of milling due to refinement in crystallite size. After 4 h of milling in alloy C, the distinct Y2O3

peaks were no longer observed, and the Ni-20Cr peaks shifted to lower diffraction angles as a result

of partial dissolution of Y2O3 in the Ni-20Cr matrix and increasing lattice parameter.

The structural quantification results are summarized in Table 2. The lattice parameters of the

fcc Ni-20Cr matrix increased after 4 h of milling indicating dissolution of Y2O3. As the milling time

increased, the average crystallite size constantly decreased from 44 nm in the blended powder to 14

nm after 2 h of milling, and thereafter decreased to 4 nm with further milling up to 4 h.

The microstructure of the as-milled Ni-20Cr-1.2Y2O3 powder after 2 h milling is shown in

Fig. 4 and exhibited a homogeneous distribution of Y2O3 particles with reduced diameter of 3 nm in

Ni-20Cr nanocrystalline matrix with reduced average size of 25 nm. The selected area diffraction

(SAD) pattern shown in Fig. 4 revealed the presence of a cubic crystal structure of Y2O3 particles

distributed in Ni-20Cr matrix with fcc crystal structure.

The density and microhardness of the Ni-20Cr-1.2Y2O3 milled for 0 h, 2 h and 4 h and then

sintered at 1100 °C for 30 min were measured and summarized in Table 3. Full density (100%) was

achieved after sintering of Ni-20Cr-1.2Y2O3 powder milled for 0 h and 4 h; however the relative

density of 2 h milled and sintered Ni-20Cr-1.2Y2O3 powder (alloy B) showed a slight reduction in

density (99.55%). The microhardness values of sintered Ni-20Cr-1.2Y2O3 alloys increased with

increasing milling time up to 4 h. This hardness increase due to longer milling time are likely due to

the refined crystallite size, accumulation of work hardening and formation of complex Y-Cr-O

particles specifically after 4 h of milling as will be shown later.

The microstructure of sintered Ni-20Cr-1.2Y2O3 alloys milled for 0 h, 2 h and 4 h (alloy A,

B and C, respectively) are shown in Figs. 5a-d, respectively. The EBSD micrograph is shown in

Fig. 5a because the grain size was too large for any TEM micrograph to be presented. That is why

EBSD micrograph was replaced with TEM micrograph in Fig. 5a. The grains in Fig. 5a were

equiaxed and fully recrystallized with average size of 8 μm. A high volume fraction of annealing

twins and Σ3 boundaries were also observed in Fig. 5a. Figure 5b shows a bright field TEM image

from the 2 h milled Ni-20Cr-1.2Y2O3 (alloy B) with randomly oriented nanograins smaller than 300

nm. The presence of homogeneously distributed nanoparticles in a large volume fraction led to a

significant grain refinement in microstructure of Ni-20Cr-1.2Y2O3 alloy. The microstructure of Ni-

20Cr-1.2Y2O3 alloy after milling for 4 h (alloy C) revealed a bimodal grain size distribution


8

containing nanograins with an average size of 120 nm as shown in Fig. 5c and coarser grains with

an average size of 400 nm as shown in Fig. 5d. The corresponding SAD patterns obtained from Fig.

5c and Fig. 5d also confirmed this bimodal grain size distribution.

High fraction of straight and annealing twin boundaries were found in all of the sintered Ni-

20Cr-1.2Y2O3 alloys regardless of milling time (alloys A, B and C) as shown in Fig. 5a and Figs.

6a-b, respectively. According to studies, formation of these twin boundaries would significantly

reduce the grain boundary energy Ni-20Cr-1.2Y2O3 powder during sintering stage [29-31].

These twin boundaries were observed in Ni-20Cr-1.2Y2O3 alloy milled for different times (0

h, 2 h and 4 h) implying that formation of these twin boundaries was due to sintering and is not

dependent on milling stage. However, the twin width in the sintered Ni-20Cr-1.2Y2O3 alloys was

strongly dependent on milling time because narrower twins were observed for alloy milled for

longer hours (4 h or alloy C) as shown in Fig. 6c. The average twin boundary spacing reduced with

increasing milling time and was measured to be 2.6 μm in alloy A, 127 nm in alloy B and 32 nm in

alloy C.

Figures 7a-b show the second phase particles in the sintered Ni-20Cr-1.2Y2O3 alloys after

milling for 2 and 4h, respectively. For alloy B, the smallest and the average particle diameter were

determined to be 3 nm and 14 nm, respectively as shown in Fig. 7a. In alloy B, three main

categories of oxide particles were found based on energy dispersive spectroscopy (EDS): (1) Ni-

based oxides in the range of 80-100 nm; (2) Cr-based oxides in the range of 20-60 nm; and (3) Y-

based oxides smaller than < 15 nm.

Similarly, the smallest and the average particle diameter were determined to be 2 nm and 7

nm, respectively for alloy C, as shown in Fig. 7b. The average particle size decreased at a longer

milling time. The particles in Fig. 7a were mainly Cr-based oxides or Y2O3 whereas the majority of

particles shown in Fig. 7b had chemical composition close to YCrO3. This could be attributed to

possible dissolution and decomposition of Y2O3 and formation of YCrO3 like new compound.

3.2. Effect of SPS parameters

The Ni-20Cr-1.2Y2O3 alloy, milled for 2 h was sintered at 600 °C for 5 min (alloy D), at

900 °C for 5 min (alloy E), at 1000 °C for 5 min (alloy F), at 1100 °C for 5 min (alloy G) and at

1100 °C for 30 min (alloy B). The physical density, relative density and microhardness values of

these alloys were measured and summarized in Table 4. The density values significantly increased

from 72.2% after sintering at 600 °C to 99.3% after sintering at 1000 °C and only slightly increased

to 99.5% and 99.6% after sintering at 1100 °C for 5 and 30 min, respectively. The main

densification occurred at 1000 °C and density values did not significantly improve with further

increase of sintering temperature or time.


9

The microhardness values constantly increased from 130.8±31.5 HV after sintering at

600 °C to 556±5 HV after sintering at 1000 °C at a constant dwell time (5 min). However, the

microhardness decreased to 470±8 HV after sintering at 1100 °C for 5 min and did not change

significantly after a dwell time of 30 min.

The overview of microstructure of Ni-20Cr-1.2Y2O3 alloy sintered at 900 °C (alloy E),

1100 °C for 5 min (alloy G) and 1100 °C for 30 min (alloy B) are shown in Figs. 8a-c, respectively.

Subgrains with an average size of 200 nm were distinguished from each other with arrays of

dislocations as shown in Fig. 8a. The grains shown in Fig. 8b were larger with an average size of

350 nm and separated from each other with well-defined sharp boundaries. After sintering of alloy

B at 1100 °C for 30 min., the average grain size was reduced even further and an extensive

dislocation activity was observed, as shown in Fig. 8c. This could be explained considering that

dynamic recrystallization phenomenon likely occurred in the alloy after 30 min at 1100 °C. During

dynamic recrystallization, dislocation activity became significant and led to more grain refinement.

Meanwhile, the interaction of nanoparticles with dislocations and mobile boundaries effectively

inhibited the grain growth process.

The twinning activity as a function of sintering temperature and time can be observed in

Figs. 9a-c. The microstructure of Ni-20Cr-1.2Y2O3 alloy sintered at 900 °C for 5 min (alloy E)

shown in Fig. 9a, contained very limited and localized twins with average width of 10 nm and

volume fraction of 2.8%. The average twin boundary spacing and volume fraction significantly

increased to 150 nm and 10%, respectively, after sintering at 1100 °C for 5 min (alloy G) as shown

in Fig. 9b. Twins in Ni-20Cr-1.2Y2O3 alloy sintered at 1100 °C for 30 min (alloy B) are shown in

Figure 9c, and the average twin boundary spacing and volume fraction were estimated to be 127 nm

and 11.6%, respectively. With increasing sintering time from 5 to 30 min, the twin boundary width

was reduced slightly, but twin boundary volume fraction increased slightly. This could be attributed

to the smaller grain size with a higher density of grain boundary areas in alloy B.

The oxide particles in alloy E and G are shown in Figs. 10a-b, respectively. The oxide

particles in alloy B were formerly shown in Fig. 7a and thus not repeated here. The average particle

size in alloy E was measured to be 4.1 nm, and the smallest particle size was found to be 2 nm.

Similarly, for alloy G, the average particle size was 12 nm, and the smallest particle size was 4 nm.

The oxide particle size increased with increasing sintering temperature and time, plausibly due to

faster kinetics of diffusion and particle coarsening at higher sintering temperature and longer

sintering time.


10

2.3. Effect of alloy composition

The physical density, relative density and microhardness values of different alloy

compositions; Ni-20Cr (alloy H), Ni-20Cr-1.2Y2O3 (alloy B) and Ni-20Cr-1.2Y2O3-5Al2O3 (alloy

I), were evaluated and are summarized in Table 5. Adding 1.2 wt.% Y2O3 to Ni-20Cr alloy

increased the relative density from 98.95±0.03% to 99.55±0.04%; however, addition of 5 wt.%

Al2O3 to Ni-20Cr-1.2Y2O3 slightly decreased the relative density to 99.18±0.02%. The

microhardness values showed significant increase from 307±3 HV in Ni-20Cr alloy to 472±8 HV in

Ni-20Cr-1.2Y2O3 and 505±10 HV in Ni-20Cr-1.2Y2O3-5Al2O3 alloy.

The microstructures of Ni-20Cr (alloy H), Ni-20Cr-1.2Y2O3 (alloy B) and Ni-20Cr-

1.2Y2O3-5Al2O3 (alloy I) are displayed in Figs. 11a-c, respectively. The microstructure of the Ni-

20Cr alloy contained fully recrystallized grains with well-defined sharp boundaries and fewer

dislocations as shown in Fig. 11a. The average grain size of the Ni-20Cr alloy was measured to be

630 nm, and the number density of oxide particles was less than that of Ni-20Cr-1.2Y2O3 earlier

shown in Fig. 7a. The effect of Y2O3 addition on the grain refinement was clearly evident in Fig.

11b. The presence of homogeneously distributed nanodispersoids in a large volume fraction led to a

higher microhardness and grain refinement in the Ni-20Cr-1.2Y2O3 alloy.

The microstructure of Ni-20Cr-1.2Y2O3-5Al2O3 (alloy I) is shown in Fig. 11c and revealed

randomly oriented grains with extensive dislocation activity and an average grain size of 385 nm.

The microstructure of Ni-20Cr alloy sintered at 1100 °C for 30 min (alloy H), contained twins with

average width of 161 nm and volume fraction of 9% as shown in Fig. 12a. The average twin width

and volume fraction in Ni-20Cr-1.2Y2O3 (alloy B) were determined to be 127 nm and 11.6%,

respectively as previously shown in Fig. 6a. The average twin width and volume fraction were

found to be 60 nm and 16.1% in the microstructure of Ni-20Cr-1.2Y2O3-5Al2O3 (alloy I) shown in

Fig. 12b. Thus, addition of 5 wt.% Al2O3 to Ni-20Cr-1.2Y2O3 alloy led to a higher volume fraction

of twins with narrower width.

The size distribution of second phase particles in Ni-20Cr (alloy H) is shown in Fig. 13a.

Coarser particles were found to be mostly located on the grain boundaries. The oxide particles in

Ni-20Cr-1.2Y2O3 (alloy B) were previously shown in Fig. 7a. A HAADF STEM micrograph

obtained from Ni-20Cr-1.2Y2O3-5Al2O3 (alloy I) is shown in Fig. 13b. The ultrafine particles with

darker contrast and smaller than 400 nm were more enriched in Al. Figure 13c revealed oxide

particles with diameter varying from 2-3 nm to 50 nm in alloy I.

The combined results of XRD, SEM/EDS and TEM/EDS revealed the presence of Cr2O3,

Cr2O5, Y2O3, Y2CrO4, and YCrO3 as well as Cr3C2 and Cr7C3 in SPSed Ni-20Cr and Ni-20Cr-

1.2Y2O3 alloys. Some of these particles were not observed in TEM due to the electropolishing

artifact. The types of particles in Ni-20Cr-1.2Y2O3-5Al2O3 (alloy I) varied between Al2O3, YAlO3


11

and negligible percentage of CrAlO3. The particle size distributions were calculated from TEM

micrographs using following equation [1];

rf

]64.1)43

[( 2/1 −=π

λ (1)

where λ is the mean particle separation (nm), r is the mean particle radius and f is the volume

fraction of dispersion particles based on several TEM micrographs. Approximately 500 particles

were considered for these calculations for better statistical data. For Ni-20Cr, the average particle

diameter, average particle radius and mean particle separation were determined to be 88 nm, 44 nm

and 187 nm, respectively. From the TEM micrographs of Ni-20Cr-1.2Y2O3 alloy, the average

particle diameter, average particle radius and mean particle separation were determined to be 14

nm, 7 nm and 58 nm, respectively. For Ni-20Cr-1.2Y2O3-5Al2O3, the average dispersion oxide

particle diameter, average particle radius and mean particle separation were determined to be 22

nm, 11 nm and 72 nm, respectively. In this calculation, the coarser particles enriched in Al were not

included.

The particle size distribution histograms for different alloy compositions are shown in Figs.

14a-c. A broad range of oxide particles were observed in Ni-20Cr alloy as shown Fig. 14a. With

addition of Y2O3 to Ni-20Cr, both average particle diameter and mean particle separation decreased

significantly. In Ni-20Cr-1.2Y2O3-5Al2O3 alloy, there was a range of nanoparticles smaller than 15

nm and coarser particles larger than 250 nm.

The compression tests were performed on Ni-20Cr, Ni-20Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-

5Al2O3 alloys at 25 °C and 800 oC by applying a strain rate of 10-3 s-1. The results are summarized

in Table 6. The true compression stress-plastic strain plots (at 800 °C) for different compositions

are illustrated in Fig. 15.

4. Discussion

4.1. Microstructural evolution

During high energy ball milling several phenomena could occur; initially Ni-20Cr powder

particles agglomerate and then continue to form fine lamellar sandwich structure until they break up

due to high energy colliding balls at later stages [25]. After milling of Ni-20Cr-1.2Y2O3 alloy for 2

h, Ni-20Cr powder became cold-welded and Y2O3 particles became finer and increased the

interfacial energy (Figs. 2c-d). As milling proceeded for 4 h, accumulative plastic strain became

significant and caused crack initiation in the milled particles as shown in Fig. 2f. After 4 h of

milling, Y2O3 particles had a sufficient driving force for decomposition. The peak broadening effect

shown in Fig. 3 can be attributed to the progressive reduction in crystallite size (i.e. grain

refinement) and increase in lattice strain arising from crystal imperfections (vacancies and


12

dislocations) and severe plastic deformation induced by high energy ball milling. No peak of Y2O3

was observed in the XRD pattern corresponding to 4 h milling and shown in Fig. 3. Dissolution of

Y2O3 in Ni-20Cr matrix was not the goal of the milling experiments; rather the intention of this

study was to homogeneously distribute Y2O3 fine particles in the Ni-20Cr matrix as shown in Figs.

4a-b. So, further milling experiments were continued up to 2 h.

In alloy A, there was no milling and plastic deformation involved prior to sintering, so large

particles and crystallites of Ni-20Cr were recrystallized by hot deformation caused during SPS.

Non-uniformly distributed Y2O3 particles were located at prior particle boundaries upon

consolidation. The recrystallized grains were dislocation free, equiaxed and without any preferred

texture.

In alloy B, recrystallization and grain growth were efficiently controlled by a high volume

fraction of oxide nanoparticles with mean particle separation of 58 nm and average radius of 7 nm

(smallest particle size of 3 nm).

With further milling for 4 h, there was a tendency for developing a bimodal grain structure

in the sintered alloy as shown in Figs. 5c-d. The reason for the bimodal grain size distribution was

likely due to heterogeneous plastic deformation and dissolution of Y2O3 in Ni-20Cr matrix after

milling for 4 h. Because the heterogeneity in Ni-20Cr powder particles incorporated by longer

milling hours from one hand and alteration in chemical compositions from one particle to another

particle or from inside to outside of a particle in another hand could initiate the bimodal grain

structure [25]. More distinct bimodal grain size distribution has already been reported for the iron-

based ODS alloys and was attributed to heterogeneity developed in the microstructure due to long

hours of milling and subsequent sintering process [17, 32, 33].

Microstructures of the sintered Ni-20Cr-1.2Y2O3 alloy milled for 0 h, 2 h and 4 h comprised

multiple twin boundaries as shown in Fig. 5a and Figs. 6a-c. With increasing milling time from 0 h

to 4 h, twin boundaries width reduced and volume fraction of twinned boundaries increased.

Presence of twin boundaries in 0 h milled sintered Ni-20Cr-1.2Y2O3 (alloy A) implied that

twin boundary formation may not depend on milling process, rather, it depended on thermal and

sintering process. Annealing twins have been frequently observed during grain growth of nickel at

temperatures of 0.68Tm and above ( i.e. ≥ 950 °C) [30], and were associated with annealing

temperature. Formation of twin boundaries in nickel alloys is because of their low stacking fault

energy. For example, adding only 0.04 mole fraction of Cr as an alloying element to pure Ni

decreases the stacking fault energy of the alloy by 40 mJ/m2 [34].

The longer milling time would produce the smaller crystallize size and consequently the

narrower twin boundaries. It is because smaller crystallite size would significantly limit the space in


13

grain interior for the twins in order to nucleate and grow [35]. Besides, crystallographic orientation

would alter more frequently in nanograins and lead to narrower twins.

Twin volume fraction was approximately determined to be 8.5%, 11.6% and 16.8% for

alloys A, B and C, respectively. The twin volume fraction increased at longer milling time. The

accurate measurement of twin volume fraction is usually challenging because some twins may not

be visible in TEM if the grains are not oriented at certain orientation or angles. Longer milling time

produced higher amount of plastic deformation and stored energy. Higher amount of stored energy

in the crystallites and crystallite boundaries of Ni-20Cr-1.2Y2O3 alloy could effectively lower the

activation energy and favor the twin boundary formation. Twin boundaries are mobile during

sintering and could minimize the grain boundary energy and total interfacial energy independently

of grain growth and grain boundary migration [30]. Therefore, higher amount of stored energy in

the milled powder can be lowered during sintering through the formation of higher volume of twin

boundaries.

The size and chemical composition of oxide particles found in the sintered Ni-20Cr-1.2Y2O3

alloys Fig. 7a-c could be different from each other as a function of milling time. It is likely that

during SPS at the elevated temperatures, Cr started precipitating out of the Ni-Cr solid solution in

the form of Cr based oxides and carbides [23]. The oxide particles of alloy B as shown in Fig. 7b

were mostly Cr-based and Y-based oxides. However, the oxide particles of alloy C shown in Fig. 7c

had composition close to Y-Cr-O. This could be due to decomposition of Y2O3 after 4 h milling as

confirmed by XRD results (shown in Fig. 3), and re-precipitation along with Cr to form smaller and

more stable particles [8]. Similar behavior was also reported in previous studies where the original

Y2O3 particles led changed into other complex more stable oxides (for example Y-Ti-O, Y-Cr-O

and Y-Al-O) [4, 36-38].

The enthalpy of formation of YCrO3 and Y2O3 are -1493 and -1907 kJ/mol, respectively

[39]. Despite the high enthalpy of formation for Y-O, the formation of Y-Cr-oxides could be

promoted due to higher concentration and diffusivity of Cr compared to Y [33, 40]. This was

provided via accumulated lattice imperfections (vacancies and dislocations) introduced by longer

milling times.

The SPS parameters such as temperature, time, heating rate and applied pressure could

significantly influence the mechanical properties and microstructural evolution. In the present

study, only the effects of SPS temperature and time were investigated on alloys D, E, F, G and B.

Densification constantly improved with increasing SPS temperature; however,

microhardness decreased at 1100 °C due to potential coarsening of oxide particles. The subgrains

formed at 900 °C were distinguished by arrays of dislocations and revealed a recovered

microstructure as shown in Fig. 8a. However, after SPS at 1000 °C, most of the grains became


14

larger, recrystallized and separated from one another with sharp boundaries as shown in Fig. 8b.

The grain size and microhardness at 1100 °C did not change with increasing time from 5 min to 30

min, suggesting that for SPS, temperature was more dominant than time.

Dislocations observed in Fig. 8c could not be due to work hardening since the hardness

values did not show any significant change between alloy G (1100 °C for 5 min) and B (1100 °C for

30 min). They could be due to dynamic recrystallization continuously occurring during SPS with

increasing time from 5 min to 30 min.

Lower hardness of alloy E compared to alloy G and B could be due to less volume fraction

of twin boundaries as shown in Figs. 9a-c. Annealing twins could contribute to hardness by

blocking dislocations and mobile grain boundaries [41]. Increasing SPS temperature increased

volume fraction of the annealing twinsalong with increase in their width. Increasing sintering time

did not significantly change twin boundary properties. Recrystallization process is controlled by

oxide dispersion and one way to facilitate the recrystallization stage is through twin formation. In

other words, higher sintering temperature may lower the activation energy of twin formation

because twinned grains can facilitate recrystallization [42].

Similar twin activities have recently been reported in powder processing and SPS of bulk

multimodal nickel [41]. Fang et al. [43] suggested that during the SPS process and applying a high

uniaxial pressure (80 MPa), certain parallel crystallographic planes might slip toward an opposite

direction maintaining a certain crystallographic distance from each other. This approach could

activate a certain twin boundary system by a critical resolved shear stress that could be achieved at

higher sintering temperatures. Randle et al. [30] suggested that twin formation kinetics in nickel

may actually be independent of grain growth and the relationship between grain growth and twin

formation may be only coincidentally related with temperature. Nonetheless, the origin of annealing

twins in Ni alloys is still a matter of conjecture.

Oxide particle size significantly increased with increasing SPS temperature as shown in

Figs. 10a-b with smallest oxide particle size found in alloy E (900 °C for 5 min). In the present

study, sintering time did not cause any significant oxide particle coarsening. Sequential adding of

Y2O3 and Al2O3 to Ni-20Cr significantly improved the hardness due to grain refinement and high

density of oxide nanoparticles as shown in Figs. 11a-c.

Fully recrystallized microstructure with less density of particles was observed in Ni-20Cr

alloy. The microstructure of Ni-20Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-5Al2O3 alloys revealed

nanograins with higher density of dislocations and oxide nanoparticles.

Higher volume fraction of second phase particles or oxide particles increased the volume

fraction of twin boundaries and reduced width of twin/matrix lamellae (Figs. 12a-b). The presence

of high volume fraction of precipitates and solutes at the boundaries of Ni-20Cr-1.2Y2O3 and Ni-


15

20Cr-1.2Y2O3-5Al2O3 alloys might impede the grain rotations and increase the activation energy of

grain rotations [44]. This could facilitate and ease formation of more annealing twins with narrower

width and higher volume fraction.

4.2. Mechanical properties and strengthening mechanisms

Higher hardness values with increasing milling time could be attributed to smaller crystallite

size, more plastic strain and narrower twin width. Because annealing twins are considered as

obstacles to dislocation motion especially when the width of twin/matrix lamellae decreases [41,

43]. Microhardness values were 307±3 HV for Ni-20Cr alloy, 471.6±7.5 HV in Ni-20Cr-1.2Y2O3

and 505±10 HV in Ni-20Cr-1.2Y2O3-5Al2O3. Park et al. [2] reported hardness values of 526.3 HV

for Ni-22Cr-11Fe-1TiO2, 645.3 HV for Ni-22Cr-11Fe-1Y2O3 and 593.3 HV for Ni-22Cr-11Fe-

0.5TiO2-0.5Y2O3 (wt.%) alloys SPSed at 1100 °C. Tang et al.[1] found adding 0.8 wt.% Hf to Ni-

0.5Al-1Y2O3 increased the hardness from 360 HV to 460 HV.

Higher hardness value in Ni-20Cr-1.2Y2O3 alloy compared to that of Ni-20Cr alloy could be

attributed to grain refinement and higher volume fraction of homogeneously distributed oxide

nanoparticles. The strengthening in Ni-20Cr-1.2Y2O3-5Al2O3 alloy could be attributed to

combination of solid solution, Hall-Petch (due to grain refinement), precipitation (due to Y2O3) and

composite strengthening (due to Al2O3 particles) mechanisms. The total strengthening was

estimated by simple linear addition of all the contributing mechanisms, assuming that strengthening

mechanisms operate independently of one another [45]. Hence, σY can be estimated by following

equation;

σY = σo + ΔσSS + ΔσD + ΔσGB + ΔσOro + ΔσC (2)

where σY is the estimated yield strength, σo is the Peierls–Nabarro stress or lattice-friction stress and

negligible, ΔσSS the solid solution strengthening contribution, ΔσD the dislocation strengthening

contribution, ΔσGB the grain boundary strengthening contribution, ΔσOro is the Orowan or dispersion

mechanism contribution and ΔσC is the composite strengthening contribution factor. The Peierls–

Nabarro or lattice-friction stress for an isotropic pure fcc crystal structure at room temperature is

generally a negligible quantity (here, a value of σo = 6-8 MPa was assumed) [46].

4.2.1. Solid solution strengthening

The solid solution strengthening due to multiple alloying elements in Ni alloys has been

investigated [47] and the strengthening can be expressed by the following relation

n

ii

niSS ck=Δ /1σ (3)

where Δσss is the solid solution contribution, ki is the strengthening constant for solute i, ci is the

concentration of solute i, and n is taken as 0.5 here. For Ni-Cr alloy, the value of k is 337 MPa at.


16

fraction-1/2 [48]. Considering the nominal composition of Ni-20Cr alloy, the strengthening due to

solutes is 158 MPa assuming that the Y2O3 and Al2O3 particles did not get dissolved in the matrix.

4.2.2. Dislocation strengthening

The contribution of dislocation strengthening can be evaluated by using Bailey–Hirsch

equation [49],

2/1ρασ GbD =Δ , (4)

where a value of 0.5 was used for the dislocation strengthening coefficient [50], and G (shear

modulus) and b (Burgers vector) were taken as 82 GPa and 0.25 nm, respectively [1, 51]. Using

several TEM micrographs, the dislocation density was calculated for the three alloys and estimated

to be 3.4×1013 m-2 for Ni-20Cr, 9.6×1013 m-2 for Ni-20Cr-1.2Y2O3, and 1.2×1014 m-2 for Ni-20Cr-

1.2Y2O3-5Al2O3. Therefore, the contributions due to dislocation strengthening were 68 MPa, 96

MPa and 111 MPa for Ni-20Cr, Ni-20Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-5Al2O3, respectively.

4.2.3. Grain boundary (Hall-Petch) strengthening

The classic Hall-Petch relationship is used for estimating the grain boundary strengthening

contribution. The average grain size for different alloys was measured and presented in the results

section. By substituting the obtained average grain sizes in the equation as suggested by Bui et al.

[52]:

ΔσGB = 5538 d-1/2, (5)

the grain boundary strengthening contributions were determined to be 257 MPa for Ni-20Cr, 360

MPa for Ni-20Cr-1.2Y2O3 and 319 MPa for Ni-20Cr-1.2Y2O3-5Al2O3 alloy.

4.2.4. Orowan (dispersion) strengthening

The additional contribution to yield strength by Orowan strengthening mechanism can be

determined by the following equation [53]:

λυπσ

)2

ln(

1

4.0 b

rGb

MOro−

=Δ (6)

where r is the average particle radius, λ the mean particle separation, f the volume fraction of

particles, υ the Poisson ratio, and M the Taylor factor. The values of λ and f were calculated to be

187 nm and 0.055 for Ni-20Cr, 58 nm and 0.235 for Ni-20Cr-1.2Y2O3, and 72 nm and 0.19 for Ni-

20Cr-1.2Y2O3-5Al2O3 alloys, respectively. In this calculation, M was taken as 3 for fcc materials

[54]. Based on these calculations, the contribution of Orowan strengthening mechanism at room

temperature was calculated to be 295 MPa for Ni-20Cr, 650 MPa for Ni-20Cr-1.2Y2O3 and 587

MPa for Ni-20Cr-1.2Y2O3-5Al2O3 alloys. The pinning effects imply that the strength of oxide

particles depends on particle size and mean particle separation. Therefore, reduction in the mean


17

particle separation and homogeneously distributed Y2O3 nanoparticles could essentially improve the

strength properties in Ni-20Cr-1.2Y2O3 compared to Ni-20Cr alloy [1].

4.2.5. Composite strengthening

Composite strengthening differs from the dispersion strengthening in terms of particle size,

which are much larger. In the particle-reinforced composite, the reinforcement phase is significantly

stronger than the matrix and during loading a large fraction of load is transferred to the

reinforcement phase. Strength of the discontinuously-reinforced composite depends on the particle

size and volume fraction. For uniformly distributed reinforcement in the matrix, the rule of mixture

can be used to calculate the net properties of the composite given by the following relation [55]:

mmppc VV σσσ += (7)

where σc is the estimated strength of the composite, Vp and Vm are the volume fraction of the

particle and matrix respectively, σp and σm are the yield strength of the particle and the matrix. Here

Al2O3 particles were added as composite reinforcement phase and its contribution was calculated to

be 295 MPa assuming the yield strength value of 2945 MPa and the volume fraction of 0.1 for the

Al2O3 particles.

Table 7 summarizes the contribution of different strengthening mechanisms in Ni-20Cr, Ni-

20Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-5Al2O3 alloys sintered at 1100 °C for 30 min and compares

the estimated yield strength values with the experimentally determined ones. The estimated yield

strength at room temperatures were found to be very close to the experimental values. For

estimating the yield strength values at 800 °C and investigating deformation mechanisms,

microstructural studies of as-compressed specimens are required that will be considered in our

future work.

The compression yield strength for the Ni-20Cr alloy was reported to be 790 MPa at 25 °C

and 125 MPa at 800 °C. It appears that the only strengthening mechanism for the Ni-20Cr alloy at

800 °C was solid solution strengthening and other mechanisms were not actively working at higher

temperatures. The compression yield values for Ni-20Cr-1.2Y2O3 alloy were reported to be 1286

MPa at 25 °C and 225 MPa at 800 °C. The yield value for INCONEL 754 (with nominal

composition of Ni-20Cr-0.3Al-0.5Ti-0.6Y2O3) was reported to be 586 MPa at 25 °C and 214 MPa

at 871 °C [56].

Figure 16 shows the variation of microhardness and true compression yield stress obtained

at room temperature (25 °C) as a function of mean particle separation (λ) for different alloy

compositions. For Ni-20Cr-1.2Y2O3 alloy, high hardness and yield stress values were obtained

through the smallest mean particle separation, and for Ni-20Cr, the lower hardness and yield stress

values were obtained due to large mean particle separation. This implies that efficiency of


18

dispersion strengthening mechanism increased with lower mean particle separation, resulting in the

improved hardness and compression yield values. In case of Ni-20Cr-1.2Y2O3-5Al2O3 alloy, high

hardness and yield stress values were obtained even though the mean particle separation was

slightly larger than that of Ni-20Cr-1.2Y2O3 alloy. This demonstrates the efficiency of Al2O3

particles as the composite reinforcement particles improving the overall hardness.

The yield strength of Ni-20Cr-1.2Y2O3 alloy was significantly higher than INCONEL 754 at

room temperature and 800°C. This suggests that strengthening mechanism at high temperatures is

due to dispersion of stable oxide particles in both alloys. Because the stable Y2O3 particles in Ni-

20Cr-1.2Y2O3 alloy not only inhibited the dislocation motion and increased the resistance of the

matrix to deformation, but also controlled the recovery and recrystallization process and inhibited

grain growth. However, for other strengthening mechanisms such as grain boundary strengthening

mechanism could be actively working along with the Orowan mechanism in the case of Ni-20Cr-

1.2Y2O3 alloy due to presence of fine grains of the alloy consolidated via SPS. This could be

regarded as an advantage of using SPS to consolidate Ni-20Cr-1.2Y2O3 alloy with finer grains. It is

worth mentioning that the contributions of fine grains and dislocation density to strength is

significant at room temperature, but those contributions would diminish significantly or disappear

altogether at high temperatures due to propensity for grain boundary sliding and reduction in

dislocation density.

5. Conclusions

In the present study, Ni-20Cr, Ni-20Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-5Al2O3 were

developed via milling and SPS. The effects of milling time, sintering time and temperature, and

alloying composition on the overall microstructure including annealing twins, particle size

distribution, as well as on the hardness and compressive yield strength of nickel based ODS alloys

were investigated. The following conclusions were drawn:

1- The grains in unmilled sintered alloys were in micron size range. The grain size was significantly

reduced after 2 and 4 h of milling. The presence or absence of milling had no influence on

manifestation of the twins. However, for longer milling times, higher amount of stored energy

decreased the activation energy of twin formation and led to a higher volume fraction of twins. The

twin width also decreased at longer milling times due to finer crystallites. Longer milling time (>4

h) led to dissolution of Y2O3 in Ni-20Cr and produced complex Y-Cr-O particles.

2- Higher sintering temperature provided higher hardness and density, and higher volume fraction

of twins. The oxide particles were stable up to 1100 °C and could efficiently control the

recrystallization process and inhibit any grain growth.

3- Adding Y2O3 to Ni-20Cr contributed to significant dispersion hardening and additional presence


19

of coarser Al2O3 provided composite strengthening to the Ni-20Cr matrix alloy. A high

compression yield stress at 800 °C was obtained for the Ni-20Cr-1.2Y2O3-5Al2O3 alloy.

Acknowledgments

The support of the University Coal Research Program of the US Department of Energy

(DOE) via a grant (DE-FE0008648) managed by the National Energy Technology Laboratory

(NETL) is gratefully acknowledged. Also, we would like to thank Professor Darryl P. Butt of the

Boise State University and the Center for Advanced Energy Studies (CAES) staff for facilitating

access to the SPS instrument

References

[1] Q. Tang, T. Hoshino, S. Ukai, B. Leng, S. Hayashi, W. Wang, Mater. Trans. 51 (2010) 2019-2024. [2] J. Park, J. Jang, T.K. Kim, S.J. Kim, J.H. Ahn, J. Nanosci. Nanotechnol. 11 (2011) 6213-6218. [3] J. Zibral, Metall. Mater. Trans. 27A (1996) 1371-1377. [4] M.K. Lee, J.J. Park, C. Rhee C., Mater. Chem. Phys. 137 (2012) 129-134. [5] D.J. Srolovitz, M.J. Luton, R.A. Petkovic, D.M. Narnett, W.D. Nix, Acta Metall. 32 (1984) 1079-

1084. [6] E. Artz, J. Rosler, Acta Metall. 36 (1988) 1053-1060. [7] B. Reppich, Acta Mater. 46 (1998) 61-67. [8] J.J. Park, H.J. Choe, S.M. Hong, M.K. Lee, C.K. Rhee, Powder Technol. 230 (2012) 139-144. [9] E. Orowan, in: Symposium on internal stresses in metals and alloys, Institute of Metals, London,

1948. p.451. [10] R.J. Arsenault, in: Marshall IH (ed.), Composite Structures 4, Springer, Netherlands, 1987, 70-81. [11] H. Wang, P. Shi, H. Yu, B. Xu, Phys. Procedia 50 (2013) 225-230. [12] A.K. Kuruvilla, V.V. Bhanuprasad, K.S. Prasad, Y.R. Mahajan, Bull. Mater. Sci. 12 (1989) 495-505. [13] E. Hornbogen, E.A. Starke, Acta Metall. Mater. 41 (1993) 1-16. [14] J. Rosler, M. Baker, Acta Mater. 48 (2000) 3553-3567. [15] C. Suryanarayana, N. Al-Aqeeli, Prog. Mater. Sci. 58 (2013) 383-502. [16] S.H. Hong, Y.G. Kim, H.Y. Kim, J.M. Poole, J.J. Debarbadillo, in: R. Darolia, J.J. Lewandowski,

C.T. Liu, P.L. Martin, D.B. Miracle, M.V. Nathal (Eds.), Proceeding of the 2nd International Conference on Structural Applications of Mechanical Alloying, Vancouver, British Columbia, 1993, p. 69.

[17] K.N. Allahar, J. Burns, B. Jaques, Y. Wu, I. Charit, J.I. Cole, D.P. Butt, J. Nucl. Mater. 443 (2013) 256-265.

[18] A.Z. Munir, U. Anselmi-Tamburini, M. Ohyanagi, J. Mater. Sci. 41 (2006) 763-777. [19] M.J. Alinger, R.G. Odette, G.E. Lucas, J. Nucl. Mater. 307 (2002) 484-489. [20] S. Kandukuri, Met. Powder Rep. 63 (2008) 22-27. [21] H. Ke, L.X. Qiang, Y. Chao, L.Y. Yuan, Trans. Nonferrous Met. Soc. China 21 (2011) 493-501. [22] J.R. Groza and A. Zavaliangos, Mater. Sci. Eng. A, 287 (2000) 171-177. [23] S. Pasebani, A.K. Dutt, I. Charit, R.S. Mishra, Mater. Sci. Forum 783-786 (2014) 1099-1104. [24] M. Nanko, M. Sato, K. Matsumaru, K. Ishikazi, Mater. Sci. Forum 510-511 (2006) 818-821. [25] B.I. Lopez, E.M. Fracnco, H. Zoz, L.G. Trapaga-Martinez, Revista Mexican De Fisica 57 (2011)

176-183. [26] C. Dong, J. Appl. Cryst. 32 (1999) 838. [27] J.B. Nelson, D.P. Riley, Proc. Phys. Soc. 57 (1945) 160-177. [28] G.K. Williamson, W.H. Hall, Acta Metall. 1 (1953) 22-31. [29] D. Horton, C.B. Thomson, V. Randle, Mater. Sci. Eng. A 203 (1995) 408-414. [30] V. Randle, P.R. Rios, Y. Hu, Scripta Mater. 58 (2008) 130-133. [31] V. Randle, The Role of the Coincident Site Lattice in Grain Boundary Engineering, The Institute of

Materials, London, 1996.


20

[32] B. Srinivasarao, K. Ohishi, T. Ohkubo, K. Hono, Acta Mater. 57 (2009) 3277-3286. [33] S. Pasebani, I. Charit, J. Alloys Compds. 599 (2014) 206-211. [34] S.L. Shang, C.L. Zacherl, Y. Fang, Y. Du, Z.K. Liu, J. Phys: Condens. Matter 24 (2012) 505403. [35] H. Wen, E.J. Lavernia, Scripta Mater. 67 (2012) 245-248. [36] M. Nganbe, M. Heilmaier, Mater. Sci. Eng. A 387-389 (2004) 609-612. [37] L. Hsiung, M. Fluss, S. Tumey, J. Kuntz, B. El-Dasher, M. Wall, B. Choi, A. Kimura, F. Willaime,

Y. Serruys, J. Nucl. Mater. 409 (2011) 72-79. [38] M.K. Miller, D.T. Hoelzer, E.A. Kenik, K.F. Russell, J. Nucl. Mater. 329-333 (2004) 338-341. [39] W. Gale, T. Totemeier, Smithells Metals Reference Book, Elsevier, Amsterdam, 2004. [40] S. Pasebani, I. Charit, Y.Q. Wu, D.P. Butt, J.I. Cole, Acta Mater. 61 (2013) 5605-5617. [41] L. Farbaniec, G. Dirras, A. Krawczynska, F. Mompiou, H. Couque, F. Naimi, F. Bernard, D.

Tingaud, Mater. Charact. 2014;doi.org/10.1016/j.matchar.2014.05.008. [42] J.L. Bair, S.L. Hatch, D.P. Field, Scripta Mater.81 (2014) 52-55. [43] S. Fang, W. Chen, Z. Fu, Mater. Design 54 (2014) 973-979. [44] C.B. Thomson, V R. Scripta Mater. 35 (1996) 385-390. [45] E. Hornbogen, P. Haasen, V. Gerold, G. Kostorz, Gupt S., Proceedings of ICAPP ’06 of the 5th

International Conference, Pergamon Press, Aachen, Germany, 1979, p.137. [46] D.M. Norfleet, D.M. Dimiduk, S.J. Polasik, M.D. Uchic, M.J. Mills, Acta Mater. 56 (2008) 2988-

3001. [47] L.A. Gypen, A. Deruyttere, J. Mater. Sci. 12 (1977) 1034-1038. [48] A. Roth, C.L. Davis, R.C. Thomson, Metall. Mater. Trans. A 28 (1997) 1329-1335. [49] J.E. Bailey, P.B. Hirsch, Philos. Mag. 5 (1960) 485-497. [50] K. Nakashima, M. Suzuki, Y. Futamura, T. Tsuchiyama, S. Takaki, Mater. Sci. Forum 503-504

(2006) 627-632. [51] T.M. Pollock, T. Sammy, J. Propul. Power 22 (2006) 361-374. [52] Q.H. Bui, S. Dirras, G. Ramtani, R. Gubicza, Mater. Sci. Eng. A 527(2010) 3227-3235. [53] L.M. Brown, R.K. Ham, in: A. Kelly, R.B. Nicholson (eds.), Strengthening Methods in Crystals,

Elsevier, Amsterdam, 1971, 9-135. [54] T.A. Parthasarathy, S.I. Rao, D.M. Dimiduk, Superalloys, TMS, 2004. [55] N. Chawla, Y. Chen, Adv. Eng. Mater. 3 (2001) 357-370. [56] C. Suryanarayana, Prog. Mater. Sci. 46 (2001) 1-184.


Fig. 1. (a) A SEM micrograp

micrograph of the Y2O3 powder

ph of the as received Al2O3 powder, and (b) a HAAADF STEM


Fig. 2. SEM micrographs of Ni-20C

h milling, (b) cross sectional observ

(d) cross sectional observation after

cross sectional observation after 4 h

Cr-1.2Y2O3 (1 wt.% stearic acid): (a) powder mor

ation after 0 h milling, (c) powder morphology a

2 h milling, (e) powder morphology after 4 h m

h milling

rphology after 0

after 2 h milling,

milling and (f)


Fig. 3. XRD patterns of Ni-20Cr-1.2

2 h (alloy B) and 4 h (alloy C)

Fig. 4. Microstructure of Ni-20Cr-1

diffraction pattern

2Y2O3 alloys milled for 0 h (alloy A),

.2Y2O3 powder milled for 2 h (alloy B) and the ccorresponding



25

Fig. 5. The micrographs of Ni-20Cr-1.2Y2O3 alloy milled for (a) 0 h (alloy A), (b) 2 h (alloy B), (c)

4 h (alloy C) showing nanograins and (d) 4 h (alloy C) showing coarse grains


26

Fig. 6. Twins in the microstructure of Ni-20Cr-1.2Y2O3 alloy milled for (a) 2 h (alloy B) and (b) 4 h

(alloy C)


27

Fig. 7. The oxide dispersoid in the microstructure of the SPSed Ni-20Cr-1.2Y2O3 alloys milled for

(a) 2 h (alloy B) and (b) 4 h (alloy C)


28

Fig. 8. The microstructure of Ni-20Cr-1.2Y2O3 alloy sintered at (a) 900 °C for 5 min (alloy E), (b)

1100 °C for 5 min (alloy G) and (c) 1100 °C for 30 min (alloy B)


29

Fig. 9. Twins in the microstructure of Ni-20Cr-1.2Y2O3 alloys sintered at (a) 900 °C for 5 min

(alloy E), (b) 1100 °C for 5 min (alloy G) and (c) 1100 °C for 30 min (alloy B)


30

Fig. 10. The oxide dispersoid in the microstructure of Ni-20Cr-1.2Y2O3 alloy sintered at (a) 900

°C for 5 min (alloy E), (b) 1100 °C for 5 min (alloy G)


31

Fig. 11. The TEM micrographs of alloys with different compositions; (a) Ni-20Cr (alloy H), (b) Ni-

20Cr-1.2Y2O3 (alloy B) and (c) Ni-20Cr-1.2Y2O3-5Al2O3 (alloy I)


32

Fig. 12. Twins in the microstructure of alloys with different compositions;

(a) Ni-20Cr (alloy H) and (b) Ni-20Cr-1.2Y2O3-5Al2O3 (alloy I) (twinning in alloy B is previously

shown in Fig. 6a)


Fig. 13. The oxide dispersoid with d

HAADF STEM micrograph obtaine

H

different compositions: (a) a BF TEM micrograp

ed from alloy I and (c) higher magnification micr

ph of alloy H, (b)

rograph of alloy


34

Fig. 14. Particle size distribution of alloys with different compositions; (a) alloy H, (b) alloy B and

(c) alloy I (approximately 500 particles were counted for each plot)


Fig. 15. True stress-true plastic strai

Ni-20Cr-1.2Y2O3-5Al2O3 alloys sint

in curves obtained at 800 °C for Ni-20Cr, Ni-20C

tered at 1100 oC for 30 min

Cr-1.2Y2O3 and


Fig. 16. The variation of microhardn

temperature) as a function of mean p

ness and true compression yield strength values

particle separation (λ).

(at room


37

Table 1. Different milling time, SPS temperatures, SPS dwell times and alloy compositions altered in processing Ni-

based ODS alloys considered for current study

Alloy (wt.%) Milling Time (h) SPS Parameters Alloy Code

Effect of Milling Time Ni-20Cr-1.2Y2O3 0 1100 oC/ 30 min A

Ni-20Cr-1.2Y2O3 2 1100 oC/ 30 min B

Ni-20Cr-1.2Y2O3 4 1100 oC/ 30 min C

Effect of SPS Parameters Ni-20Cr-1.2Y2O3 2 600 oC/ 5 min D

Ni-20Cr-1.2Y2O3 2 900 oC/ 5 min E

Ni-20Cr-1.2Y2O3 2 1000 oC/ 5 min F

Ni-20Cr-1.2Y2O3 2 1100 oC/ 5 min G

Ni-20Cr-1.2Y2O3 2 1100 oC/ 30 min B

Effect of Alloy Composition Ni-20Cr 2 1100 oC/ 30 min H

Ni-20Cr-1.2Y2O3 2 1100 oC/ 30 min B

Ni-20Cr-1.2Y2O3-5Al2O 2.5 1100 oC/ 30 min I

Table 2. Microstructural parameters of Ni-20Cr-1.2Y2O3 powder

milled for 0 h, 2 h and 4 h as determined by XRD

Milling Time

(h)

Crystallite Size

(nm)

Lattice Strain

(%)

Lattice Constant

(nm)

0 44±12 0.03±0.001 0.3530±0.0002

2 14±7 0.03±0.001 0.3536±0.0003

4 4±2 0.15±0.003 0.3560±0.0004

Table 3. Physical density, relative density and microhardness values using different milling times; for alloys A, B and C

Alloy Physical Density

(g/cm3)

Relative Density

(%)

Microhardness

(HV)

A (0 h) 8.228±0.005 ~100 202±6

B (2 h) 8.170±0.008 99.55±0.10 472±7

C (4 h) 8.226±0.005 ~100 527±21


38

Table 4. Physical density, relative density and microhardness values of Ni-20Cr-1.2Y2O3 alloys sintered at different

temperatures and times; for alloys D, E, F, G and B


(g/cm3)

Relative Density

(%)

Microhardness

(HV)

D (600 °C for 5 min) 5.924±0.020 72.18±0.24 131±31

E (900 °C for 5 min) 7.708±0.002 93.93±0.02 395±11

F (1000 °C for 5 min) 8.150±0.030 99.26±0.30 556±4

G (1100 °C for 5 min) 8.163±0.004 99.48±0.04 470±7

B (1100 °C for 30 min) 8.170±0.008 99.55±0.10 472±7

Table 5. Physical density, relative density and microhardness values using different alloying composition; for alloys H,

B and I


(g/cm3)

Relative Density

(%)

Microhardness

(HV)

H (Ni-20Cr) 8.19±0.01 98.95±0.03 307±3

B (Ni-20Cr-1.2Y2O3) 8.17±0.01 99.55±0.04 472±7

I (Ni-20Cr-1.2Y2O3-5Al2O3) 7.70±0.01 99.18±0.02 505±10

Table 6. The compression yield stress values of Ni-20Cr, Ni20-Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-5Al2O3 alloys

sintered at 1100 °C for 30 min

Alloy Compression Yield

Stress at 25 °C (MPa)

Compression Yield

Stress at 800 °C (MPa)

H (Ni-20Cr) 790 125

B (Ni-20Cr-1.2Y2O3) 1286 225

I (Ni20-Cr-1.2Y2O3-5Al2O3) 1470 250

Table 7. Contribution of different strengthening mechanisms in Ni-20Cr, Ni20-Cr-1.2Y2O3 and Ni-20Cr-1.2Y2O3-

5Al2O3 sintered at 1100 °C for 30 min

Alloy σo (MPa) ΔσSS

(MPa)

ΔσD

(MPa)

ΔσGB

(MPa)

ΔσOro

(MPa)

ΔσC

(MPa)

σY Calculated

(MPa)

σY

Experimental

(MPa)

H (Ni-20Cr) 8 158 68 257 295 0 786 790

B (Ni-20Cr-1.2Y2O3) 8 158 96 359 650 0 1271 1286

I (Ni20-Cr-1.2Y2O3-5Al2O3) 8 158 111 319 587 295 1478 1470


Date post:	23-Jan-2022
Category:	Documents
Upload:	others
View:	5 times
Download:	0 times

Oxide Dispersion Strengthened Nickel Based Alloys via ...

Documents